The timely detection and elimination of bridge health risks are related to urban transportation and people’s travel safety. In this paper, Bragg grating (BG) fiber optic sensors are selected to scan the light source and modulate the reflected signal of the bridge structure to obtain the bridge health monitoring signal source. The bridge model is constructed based on the Hilbert-Huang transform (HHT) method to identify the parameters such as the intrinsic frequency and damping ratio of the bridge structure under different degrees of freedom systems. The Wolf Pack Algorithm (WPA) is introduced to accelerate the solving speed of the optimal solution of the hierarchical weights of the BP neural network, which improves the instantaneous efficiency of the bridge health risk monitoring. The BG fiber optic sensors have better performance than the same type of sensors in the three categories of anti-electromagnetic interference, anti-drift, and anti-reciprocal anthropomorphic static force, and are able to collect the accurate bridge health data and improve the accuracy of the bridge model construction. Using the wolf pack algorithm to optimize the BP neural network, the predicted probability of risk assessment for 50 bridge parts is 0.091~0.989, which is very close to the actual 0.043~0.979.
